Mislocalization of Fukutin Protein by Disease-causing Missense Mutations Can Be Rescued with Treatments Directed at Folding Amelioration*

Background: The molecular pathogenesis of fukutin-deficient dystroglycanopathy remains unclear, and no effective treatment is available. Results: Some disease-causing missense fukutin mutants showed mislocalization in cultured cells, which can be corrected by treatments directed at folding amelioration. Conclusion: Correction of cellular localization of disease-causing mutants may have a therapeutic benefit. Significance: A possible therapeutic strategy for fukutin-deficient dystroglycanopathy is proposed based on its molecular pathogenesis. Fukuyama-type congenital muscular dystrophy (FCMD), the second most common childhood muscular dystrophy in Japan, is caused by alterations in the fukutin gene. Mutations in fukutin cause abnormal glycosylation of α-dystroglycan, a cell surface laminin receptor; however, the exact function and pathophysiological role of fukutin are unclear. Although the most prevalent mutation in Japan is a founder retrotransposal insertion, point mutations leading to abnormal glycosylation of α-dystroglycan have been reported, both in Japan and elsewhere. To understand better the molecular pathogenesis of fukutin-deficient muscular dystrophies, we constructed 13 disease-causing missense fukutin mutations and examined their pathological impact on cellular localization and α-dystroglycan glycosylation. When expressed in C2C12 myoblast cells, wild-type fukutin localizes to the Golgi apparatus, whereas the missense mutants A170E, H172R, H186R, and Y371C instead accumulated in the endoplasmic reticulum. Protein O-mannose β1,2-N-acetylglucosaminyltransferase 1 (POMGnT1) also mislocalizes when co-expressed with these missense mutants. The results of nocodazole and brefeldin A experiments suggested that these mutant proteins were not transported to the Golgi via the anterograde pathway. Furthermore, we found that low temperature culture or curcumin treatment corrected the subcellular location of these missense mutants. Expression studies using fukutin-null mouse embryonic stem cells showed that the activity responsible for generating the laminin-binding glycan of α-dystroglycan was retained in these mutants. Together, our results suggest that some disease-causing missense mutations cause abnormal folding and localization of fukutin protein, and therefore we propose that folding amelioration directed at correcting the cellular localization may provide a therapeutic benefit to glycosylation-deficient muscular dystrophies.

Fukuyama-type congenital muscular dystrophy (FCMD, 2 MIM 253800) is the second most common childhood muscular dystrophy and one of the most prevalent autosomal recessive disorders in the Japanese population. FCMD is clinically characterized by congenital muscular dystrophy in combination with cortical dysgenesis (micropolygyria) and ocular abnormalities (1). We identified fukutin, the gene responsible for FCMD, on chromosome 9q31 by linkage analysis and positional cloning (2,3). Most FCMD-bearing chromosomes have been derived from a single ancestral founder and have a 3-kb retrotransposal insertion in the 3Ј noncoding region of the fukutin gene. Compound heterozygosity, with both a retrotransposal mutation and a point mutation in fukutin, is sometimes seen and generally exhibits more severe pathologies (4,5). However, a recent report has identified several Japanese patients presenting with mild limb-girdle dystrophy (LGMD2M, MIM 611588) and normal intelligence (6) and who have a retrotransposal mutation and a point mutation in the fukutin gene. Outside Japan, fukutin mutations have been reported in patients with various phenotypes, from Walker-Warburg syndrome (WWS, MIM 236670) to LGMD (7)(8)(9)(10)(11)(12)(13). Overall, the current common understanding is that fukutin alterations can give rise to a wide spectrum of phenotypes.
Mutations in fukutin cause abnormal glycosylation of the cell surface laminin receptor ␣-dystroglycan (DG) and reduce its laminin binding activity (14). The ␣and ␤-DG complex is believed to provide physical strength to the sarcolemma by connecting the basal lamina to the cytoskeleton. Thus, abnormal glycosylation caused by fukutin mutations underlies FCMD molecular pathogenesis, but the exact function of fukutin remains unclear. The fukutin gene encodes a 461-amino acid protein with a predicted molecular mass of 53.7 kDa (3). Although endogenous fukutin protein has not been detected in cells, likely due to its low abundance, expression studies have proposed that fukutin gene product localizes to the Golgi apparatus (15,16). Fukutin protein contains a transmembrane domain (3,16), a putative N-glycosylation site (3), and a DxD motif that is predicted to modify cell surface glycoproteins or glycolipids (17). Previously, we showed that the transmembrane domain of fukutin binds to the protein O-mannose ␤1,2-N-acetylglucosaminyltransferase 1 (POMGnT1), which is encoded by the responsible gene for muscle-eye-brain disease (MIM 253280) (18), suggesting that fukutin affects the enzymatic activity of POMGnT1 (16). Other regions of the fukutin protein share no sequence homology with known proteins. In addition to FCMD, several other forms of muscular dystrophy are caused by abnormal glycosylation of ␣-DG; together, these conditions are termed as "dystroglycanopathy." To date, six genes (protein O-mannosyltransferase 1 (POMT1), protein O-mannosyltransferase 2 (POMT2), POMGnT1, fukutin, fukutin-related protein (FKRP), and LARGE) have been implicated in dystroglycanopathies, and all are thought to be involved in glycosylation of ␣-DG (19 -23). POMGnT1 and the POMT1/2 complex are known to have glycosyltransferase activities directly involved in synthesis of O-mannosyl sugar chains on ␣-DG (18,24). Quite recently, it has been shown that LARGE can act as a bifunctional glycosyltransferase with both xylosyltransferase and glucuronyltransferase activities (25). On the other hand, the exact function of FKRP is unknown. Yoshida-Moriguchi et al. reported that a phosphodiester-linked moiety on O-mannose of ␣-DG is defective in LARGE-or fukutin-deficient dystroglycanopathies (26). This finding suggests that LARGE and fukutin might be involved in the synthesis of the postphosphoryl modification, which is necessary for laminin binding activity.
The precise pathogenic mechanism of FCMD has remained obscure. In this report, to understand molecular pathogenesis of fukutin-deficient muscular dystrophies, we constructed 13 disease-causing missense fukutin mutations that have been reported inside and outside Japan (4 -6, 9 -13, 27, 28) and investigated their pathological roles in fukutin intracellular location. Four mutants (A170E, H172R, H186R, and Y371C) lost their Golgi localization and instead accumulated in the endoplasmic reticulum (ER) when expressed in C2C12 cultured cells. Using fukutin-null mouse embryonic stem (ES) cells, we showed that these mutants retain the activity responsible for ␣-DG glycosylation. Finally, we found that low temperature culture and curcumin treatment are effective in correcting the localization of these missense fukutin mutants.

EXPERIMENTAL PROCEDURES
Reagents-Brefeldin A (BFA) and nocodazole were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Curcumin was purchased from Nacalai Tesque (Kyoto, Japan).
Cell transfection was performed using Effectene (Qiagen) according to the manufacturer's protocol. 48 h after transfection, the cells were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) and then permeabilized in PBS with 0.5% Triton X-100 (Nacalai Tesque). For immunofluorescence detection, after blocking with 1% BSA (Wako) in PBS at room temperature for 1 h, the cells were first incubated for 90 min with polyclonal anti-FLAG, monoclonal anti-GM130, monoclonal anti-KDEL, or monoclonal anti-V5 antibodies, followed by Alexa Fluor 488-conjugated anti-rabbit IgG and/or Alexa Fluor 546-conjugated anti-mouse IgG (Invitrogen) for 1 h at room temperature. After a final rinse with PBS, cells were observed by fluorescence microscopy using a Leica DMR microscope (Leica Microsystems, Wetzlar, Germany).
For BFA, nocodazole, or curcumin treatment, the cells were incubated with 5 g/ml BFA or 10 g/ml nocodazole for 2 h after 48 h of transfection, or 10 g/ml curcumin for 24 h after 24 h of transfection. For statistical analysis of fukutin cellular localization, cells expressing fukutin were classified into four classes (Golgi localization, Golgi and around localization, dot localization, and ER localization) (see Fig. 6A). For statistical analysis of POMGnT1 localization, cells co-expressing fukutin/ POMGnT1 were classified into three classes (Golgi localization with fukutin, Golgi localization without fukutin, and ER localization). The number of cells in each class was counted and analyzed using the 2 test.
Dystroglycan Preparation-DG was enriched from solubilized mouse ES cells. The ES cells were solubilized in 1 ml of PBS containing 1% Triton X-100 and protease inhibitor mixture (Nacalai Tesque). Solubilized fractions were incubated with 30 l of wheat germ agglutinin-agarose beads (Vector Laboratories) at 4°C for 2 h. The beads were washed five times with 1 ml of PBS containing 0.1% Triton X-100 and protease inhibitor mixture, then directly boiled for 5 min in SDS-polyacrylamide gel loading buffer.
SDS-PAGE, Western Blotting, and Laminin Overlay Assay-Cell lysates were dissolved in SDS sample buffer and subjected to SDS-PAGE in 10% gels or 7.5% gels. Gels were transferred to polyvinylidene difluoride (PVDF) membranes (Millipore), probed with anti-FLAG, anti-␣-DG core protein (AP-074G-C), anti-␣-DG sugar chain (IIH6C4), or anti-␤-DG antibodies, and then developed with horseradish peroxidase-conjugated secondary antibody (DAKO, Carpinteria, CA). Blots were processed using ECL plus Western blotting detection system (GE Healthcare) and exposed to Fuji RX-U x-ray film (Fuji Film, Kanagawa, Japan). The laminin overlay assay was performed according to the method of Michele et al. (14).
Additional methods are described under supplemental Experimental Procedure.

Mislocalization of Disease-causing Missense Mutant Fukutin
Proteins-To examine the cellular location of fukutin proteins containing disease-causing missense mutations, we constructed expression vectors encoding wild-type or mutant fukutin proteins with a FLAG epitope at the C terminus. The missense mutants analyzed in this study have been identified inside and outside Japan (Table 1), and their clinical phenotypes vary from severe WWS-like to mild LGMD-type without mental retardation (Table 1 and Fig. 1). These constructs were transfected into C2C12 myoblast cells, and the cellular localizations of the expressed fukutin proteins were examined by immunofluorescence. Immunofluorescent signals indicated co-localization of the expressed wild-type fukutin with the Golgi apparatus marker GM130 (162/176 cells; Golgi ϩ Golgi and around/total cells) ( Fig. 2A, merge, and supplemental Table  I Table I), instead showing co-localization with the ER marker KDEL (Fig.  2B). These results indicated that A170E, H172R, H186R, and Y371C aberrantly localize to the ER.
Accumulation of Missense Fukutin Mutants in ER Caused by Impaired Transport to the Golgi-Accumulation of the four mutants in the ER might result from improper cellular trafficking. To determine whether the mutants are not properly transported from the ER to the Golgi or whether they are transported back to the ER after reaching the Golgi, we treated C2C12 cells expressing wild-type fukutin or the four mutants with nocodazole (an inhibitor for retrograde transport from the Golgi to the ER) or BFA (an inhibitor for anterograde transport from the ER to the Golgi). If the four mutants reached the Golgi and then were immediately transported back to the ER, the four mutant proteins should be detected in the Golgi after nocodazole treatment. Immunofluorescent signals indicating ER accumulation of the four mutants were observed following a 2-h incubation with 10 g/ml nocodazole ( Fig.  3C-F, panels 4). When cells expressing wild-type fukutin were incubated with 5 g/ml BFA, the wild-type fukutin was detected in the ER (99/107; ER/total cells) (Fig. 3B, panel 6, and supplemental Table II), as seen in the cells expressing any of the four mutants. These data suggested that failure of proper transport via the anterograde pathway causes mislocalization of the four mutants to the ER.
Correction of Cellular Location of Mutant Fukutin Proteins by Low Temperature Culture-We hypothesized that mislocalization resulted from protein misfolding and therefore examined whether the localization of the four missense mutants could be corrected by folding amelioration. It has been reported that cell culture at low temperature can ameliorate folding and correct the subcellular localization of missense mutant proteins (31). In culture at 37°C, the four missense fukutin mutants (A170E, H172R, H186R, and Y371C) were co-localized with KDEL ( Fig. 4A). At low temperature (27.5°C), the A170E (130/ 144; Golgi ϩ Golgi and around/total cells), H172R (130/159), and Y371C (102/158) mutants preferentially co-localized with GM130 ( Fig. 4A and supplemental Table I), indicating that the ER accumulation had decreased and proper localization to the Golgi was restored. Most of the expressed H186R mutant protein, however, remained in the ER (134/155; ER/total cells) at 27.5°C; only a small proportion of this mutant shifted to the Golgi ( Fig. 4A and supplemental Table I).

Correcting Disease-causing Fukutin Mutant Mislocalization
the ER (108/141; Golgi without fukutin/total cells) ( Fig. 4B and supplemental Table III). These data showed that low temperature culture could correct mislocalization of the mutant fukutin proteins, but its effect may depend on the position of missense mutations within the protein.

Residual Function of Fukutin Missense Mutants to
Restore ␣-Dystroglycan Glycosylation-C2C12 cells contain endogenous fukutin, making it difficult to evaluate the ␣-DG modification activity of exogenously expressed mutant fukutin proteins. Instead, we used fukutin-null mouse ES cells (FCMDp182 cell) (30) to examine whether the four missense mutant fukutin proteins retain the activity responsible for ␣-DG modification. We expressed the four mutants (A170E, H172R, H186R, and Y371C) in FCMDp182 cells (supplemental Fig. 1) and then examined the recovery of glycosylation and laminin binding activity of ␣-DG. In mock-transfected FCMDp182 cells cultured at either 37°C or 27.5°C, ␣-DG showed no detectable reactivity against the monoclonal IIH6C4 antibody, which recognizes the functionally glycosylated form of ␣-DG. We also observed hypoglycosylation of ␣-DG, as indicated by lower molecular mass, and no ␣-DG laminin binding activity in mock-transfected cells (Fig. 5, lanes 2 and 10). When wild-type fukutin was expressed in FCMDp182 cells, we observed restoration of the laminin binding activity and the IIH6C4 reactivity of ␣-DG, but at much weaker levels than for ␣-DG in the wildtype mouse ES cell line AB2.2 (lanes 1 and 3). This partial restoration and the faint amount of ␣-DG core protein bands around 100 kDa (the size of ␣-DG in the AB2.2 cells) may result from low transfection efficiency. Expression of each of the four missense mutants in the FCMDp182 cells restored the IIH6C4 reactivity and the laminin binding activity to levels comparable with those observed in the wild-type fukutin transfectants (lanes 3-7). Although these missense mutants preferentially localized to the ER at 37°C, it is possible that small amounts of the expressed mutants reached the Golgi. To support this interpretation, a report has shown that the mislocalized mutant protein ⌬F508 cystic fibrosis transmembrane conductance regulator, which is the most common mutation leading to cystic fibrosis, shows partial function when overexpressed (32). We have not detected restoration of the IIH6C4 reactivity and the laminin binding activity by expression of a mutant fukutin protein with direct substitutions in the DxD motif (D317A/V/ D319A) (supplemental Fig. 2). Expression of FKRP, a member  No co-localization was seen for the missense fukutin mutants A170E, H172R, H186R, and Y371C. Wild-type (WT fukutin) and the other nine mutants co-localized with GM130. B, mislocalization of missense fukutin mutants to the ER. The fukutin missense mutants A170E, H172R, H186R, and Y371C (green) lost their localization to the Golgi (red) in transfected C2C12 cultured cells (upper). Co-localization of these mutants with the ER marker KDEL (red) was seen in the cultures (lower). Blue, DAPI, indicating the nucleus. Scale bars, 10 m. of the fukutin protein family, in FCMDp182 cells showed no effect on ␣-DG glycosylation (lane 8). We also performed these experiments at low culture temperature (27.5°C) and obtained similar results (Fig. 5, lanes 9 -16). The reason for reduced efficiency of glycosylation recovery in low temperature culture is unclear, but the low temperature might affect enzymatic activities involved in the ␣-DG glycosylation pathway. Because the expression level of each mutant protein was different (Fig. 5, anti-FLAG), it was not possible to simply compare their residual activity. Importantly, however, all four mutants retained the activity responsible for ␣-DG modification.
Correction of Mislocalization of Missense Fukutin Mutants by Curcumin Treatment-Our results indicated that mislocalization of some missense mutants could cause disease-related abnormal glycosylation of ␣-DG. Therefore, we next searched for chemicals that could restore proper localization. Curcumin, a nontoxic natural constituent of turmeric spice, has shown the ability to correct misfolding and mislocalization of the ⌬F508 mutant (33). We incubated C2C12 cells expressing the four mutants (A170E, H172R, H186R, or Y371C) with 10 g/ml curcumin at 37°C for 24 h. Among the four mutants, the A170E mutant showed the greatest benefit from curcumin treatment. In the absence of curcumin, approximately 90% of the cells showed ER localization of the A170E mutant (159/175; ER/total cells), and only a few cells (7/175; Golgi ϩ Golgi and around cells) showed Golgi localization (Fig. 6). Curcumin treatment significantly decreased the ER mislocalization signal of the A170E mutant (69/156; ER/total ells) and increased the Golgiand Golgi/around signals (72/156; Golgi ϩ Golgi and around cells) (p ϭ 8.28 ϫ 10 Ϫ20 ). Although not as dramatic as seen with A170E, the other mutants showed slight beneficial changes in their cellular distributions following curcumin treatment ( Fig.  6B and supplemental Table II). We also examined glycerol, arginine, and 17-allylaminogeldanamycin for beneficial effects on localization of the mutants, but none was observed (data not shown).

DISCUSSION
In this report, we have demonstrated that some disease-causing missense fukutin mutants lost their Golgi localization in C2C12 cultured cells and that this mislocalization can be corrected by low temperature culture or curcumin treatment. We identified four missense mutants that localized abnormally to the ER. Nocodazole treatment did not alter their ER localization, and low temperature culture shifted three of the four mutants to the Golgi. It is generally recognized that cell culture at low temperature can ameliorate protein misfolding and correct abnormal cellular localization (31). Therefore, we presume that some missense mutants could not be transported to the Golgi via the anterograde pathway because of protein misfold- ing caused by amino acid substitutions. A large amount of the fourth mutant protein, H186R, retained the ER localization even under low temperature conditions (Fig. 4A). It has been reported that the patient with the homozygous H186R mutation is affected with WWS (12), which shows more severe pathological features than typical FCMD. The H186R mutation in the fukutin gene may affect protein folding severely enough that low temperature conditions cannot correct mislocalization. We reported previously that POMGnT1 interacts with fukutin and co-localizes to the Golgi (16). Our present data show that correction of the mislocalization of the three missense fukutin mutants by low temperature culture was accompanied by correction of POMGnT1 cellular localization (Fig. 4B). This POMGnT1 behavior is rational because the transmembrane region of fukutin, through which fukutin binds to POMGnT1 (16), remained intact in these missense fukutin mutants (Fig. 1,  single underline). Immunoprecipitation experiments confirmed the interaction between each of the missense fukutin mutants and POMGnT1 (supplemental Fig. 3). Most of the H186R mutant remained in the ER at 27.5°C, but POMGnT1 expressed with the H186R localized to the Golgi. The reason for this result is uncertain, but a possible explanation is that the H186R mutant may have a harmful (dominant negative-like) effect on POMGnT1 localization, and this effect may be ameliorated at low temperature.
The above-mentioned four mutants, which showed abnormal localization to the ER, were identified in patients presenting with a typical or a severe phenotype. However, several of the remaining nine mutants, which showed Golgi localization, were also identified in patients presenting with the typical or the severe phenotype (G125S, M133T, C250G, and W305C) ( Table 1). Therefore, the typical or severe phenotypes seem not always to be related to abnormal cellular localization of mutated proteins. Given that FCMD is inherited in an autosomal recessive manner, disease-causing mutations in the fukutin gene must lead to loss of function of the fukutin protein.
Although the exact function of fukutin is undetermined, these nine mutations may disrupt an important functional domain in the protein. For example, the W305C (10) and R307Q (9,13) mutations are located in the DxD motif (Fig. 1, double underline), which is predicted to be involved in the modification of cell surface glycoproteins or glycolipids (17). These substitutions may disrupt the DxD motif and produce dysfunctional fukutin protein. Five mutations (A114T, R179T, R246G, R307Q, and Q358P) have been identified in patients presenting with mild phenotypes (congenital muscular dystrophy with no mental retardation, LGMD with no mental retardation, or cardiomyopathy) (6,9,10,13). In the present study, these five mutant proteins were localized to the Golgi when expressed in C2C12 cells. In patients with R179T or Q358P mutations, ␣-DG shows residual reactivity against the monoclonal antibody IIH6C4, which recognizes functionally glycosylated ␣-DG (6), indicating that these fukutin mutations retain partial function in the DG maturation pathway.
It is of interest that some missense fukutin mutants retain ␣-DG glycosylation activity and that their mislocalization could be partly corrected by treatments directing at folding amelioration. These observations suggest that drugs capable of correcting the localization might have therapeutic benefits in patients who carry these missense fukutin mutants. Although approximately half of the cells expressing the A170E mutant retained the ER accumulation signals following curcumin treatment (Fig. 6), recent studies have indicated that even partial restoration of ␣-DG glycosylation can produce therapeutic effects (29). Efforts to identify more efficient folding amelioration reagents may lead to therapeutic strategies.
A large number of missense mutations have been identified in dystroglycanopathy. It has been reported that missense mutations in POMT1 and POMGnT1 compromise enzymatic activity in the gene products (18,34,35). Disease-causing missense mutations in POMT2, LARGE, and FKRP have been also identified (10, 21-23, 36, 37). Recently, Kawahara et al. have reported that expression of some disease-causing missense FKRP mutant proteins in FKRP knock-down zebrafish restores ␣-DG glycosylation, suggesting a residual FKRP function in these missense mutants (38); however, phenotype improvement depends on the location of the mutation. The FKRP missense mutations C318Y and A455D, which failed to improve the fish phenotype in the report from Kawahara et al. (38), were reported to show abnormal cellular localization when expressed in certain cell lines (39). On the other hand, using several different cell lines, Dolatshad et al. suggested that a reduced protein (putative enzymatic) function of FKRP rather than protein mislocalization is the primary mechanism of disease (39). Interestingly, Bao et al. have reported cells lacking FKRP transcripts but expressing IIH6C4-reactive ␣-DG (40). This finding may indicate a possibility of a FKRP-independent glycosylation pathway. Alternatively, only a subtle amount of FKRP, even below detectable level by RT-PCR, may be sufficient for ␣-DG glycosylation. This implies that some missense mutants can restore IIH6C4-reactive ␣-DG if only a little protein function remains. Of the increasing number of identified disease-causing missense mutations, some may alter the cellular location of the protein, which can be a direct cause of disease. Correction of cellular localization or folding amelioration may have a therapeutic benefit for dystroglycanopathies caused by missense mutations, although the finding from cell culture experiments must be interpreted cautiously when extrapolating to human disease. FIGURE 5. Fukutin missense mutants retain the activity responsible for ␣-DG glycosylation. Western blot analysis was performed to detect IIH6C4 (glycosylated ␣-DG), core ␣-DG, ␤-DG, and FLAG tag, and the laminin overlay assay. DG and fukutin proteins were prepared from wild-type mouse ES cells (AB2.2), fukutin-deficient mouse ES cells (FCMDp182), and fukutin-or FKRPtransfected FCMDp182 cells, and cultured at 37°C and 27.5°C. FCMDp182 cells expressing any of the mutant fukutin proteins showed levels of glycosylated ␣-DG signal (IIH6C4-reactivity) and laminin binding activity that were comparable with those observed in FCMDp182 cells expressing wild-type fukutin. FIGURE 6. Curcumin treatment partly corrects mislocalization of the missense fukutin mutants. A, effects of curcumin treatment on the cellular localization of the missense A170E fukutin mutants. C2C12 cells expressing the missense fukutin mutant A170E were cultured in the presence or absence of 10 g/ml curcumin. Cells were classified into four classes (Golgi localization, Golgi and around localization, dot localization, and ER localization), then counted and statistically analyzed using the 2 test. Red, GM130; green, expressed missense fukutin protein A170E; blue, DAPI. Scale bar, 10 m. B, statistical analysis of effects of curcumin treatment on the cellular localization of the missense fukutin mutants A170E, H172R, H186R, and Y371C.
In this study, we have used curcumin to correct mislocalization of missense fukutin mutants. It has been reported that curcumin can correct misfolding and mislocalization of cystic fibrosis transmembrane conductance regulator with the ⌬F508 mutation (33). Curcumin has also been reported to have protective effects in neurodegenerative diseases by inhibiting protein misfolding and aggregation in Creutzfeldt-Jakob disease and Parkinson disease (41,42). These studies indicate that curcumin or its derivatives might be new candidate compounds for protein-folding diseases. In addition, the use of pharmacological chaperones to stabilize or promote correct folding of mutant proteins has been shown as a potential therapeutic approach to phenylketonuria, in which more than 500 diseasecausing mutations have been identified (43). Our results contribute to a deeper understanding of the molecular pathogenesis of fukutin-deficient muscular dystrophy and have led us to propose a novel therapeutic strategy directed at correction of cellular localization and/or folding amelioration of diseasecausing missense mutant proteins.